Cold Working Process on Hard MetalStacked Assembly

Guillaume Pichon1(B), Alain Daidie1, Adeline Fau1, Clément Chirol2,and Audrey Benaben2

1 Université de Toulouse, Institut Clément Ader, UMR CNRS 5312, INSA/UPS/ISAE/MinesAlbi, Toulouse, France

guillaume.pichon@airbus.com2 Airbus Operations S.A.S., Toulouse, France

Abstract. Designed for aeronautical and automotive applications, the split sleevecold expansion process is used to improve the fatigue life of bolted metallic parts.Although its application has been well tested on aluminum assemblies, hard metalapplications are still being studied. This paper presents experimental results ofdouble bolt joint assemblies under double shear fatigue tests after stacked splitsleeve cold expansion. The behaviors of two sizes of assemblies with differentdegrees of expansion are investigated. S-N curves are the main indicators of thisstudy but thermal aspects are also investigated to observe fretting in the specimensas bolts are preloaded. Bolt tension is a major parameter in assembly regardingfatigue life. Interference between those two phenomena is at the heart of thispaper. The first results show that stacked cold expansion has a negative effecton mechanical performances, as it deteriorates the fatigue life of the assembly.However, an examination of these results provides a coherent explanation for theloss of performance that occurs.

Keywords: Cold expansion · Process · Fatigue · Assembly · Bolt

1 Introduction

Reducing the weight of structures is one of the major aims of aeronautical engineers, andspecifically concerns structural parts made of hard, heavy metal. Furthermore, modernconcern regarding environmental issues has strengthened the will of aircraft manufactur-ers to focus their innovation efforts on weight loss in order to reduce fuel consumption.For instance, the Dutch aviation industry has planned to cut its global CO2 emissions byone-third by the year 2030 [1]. Some studies focus on newmaterials such as composites,which have created a breakthrough in the field of aeronautical structures [2]. Neverthe-less, their complexity and characteristics (mechanical and thermal) make them unusablefor some structural parts so far. Hard metals, and titanium alloys in particular, remain thebest solutions in certain cases and are thus frequently used [3]. However, developing newand more efficient technology on ancient material is also a field of study that remainsa technical challenge. Aeronautical structures are mainly made of sub-assemblies held

© The Author(s) 2021L. Roucoules et al. (Eds.): JCM 2020, LNME, pp. 41–47, 2021.https://doi.org/10.1007/978-3-030-70566-4_8

42 G. Pichon et al.

together by a significant number of fasteners (rivets or bolts). The cold expansion pro-cess is an innovative way to enhance the fatigue life of metal components by reinforcingthe hole edges to delay their failure. It is then possible to reduce the design weight andachieve identical mechanical performances, while saving weight on the total structure.Several cold expansion process exist: expansion by conical pin, the British railwaysprocess, and a conical four-split mandrel. This paper focus on the split sleeve expansionprocess.

2 Stacked Cold Expansion Process

TheColdExpansion patented by FatigueTechnologies Inc. [4, 5] is themostwidely used,mainly because of its speed of execution and portability [6]. By pulling an oversizedmandrel through a fastener hole, compressive residual stresses are developed aroundthe hole wall, fatigue crack propagation is delayed and maintenance sessions can bescheduled less frequently [7]. The process is performed with the help of a split sleeve,placed longitudinally between the mandrel and the hole. A conical taper mandrel isscrewed at the extremity of a pressure cylinder. To guide the mandrel and to maintain thesleeve in position during expansion, a nosecap, symmetrically split into four sections,is inserted around the mandrel and directly mounted onto the puller unit frame. Then,the split sleeve is placed around the mandrel, and the mandrel and sleeve assembly isinserted inside the specimen hole. Finally, as the mandrel travels through the material,the hole is expanded at a ratio defined by the Degree of Cold Expansion (DCE) expressedby Eq. (1)

DCE = dm0 + 2tss − dsp

0

dsp0

× 100 (1)

where d0m is the mandrel major diameter, tss is the split sleeve thickness and d0sp is theinitial hole diameter.

In this particular study, the mandrel is extracted by travelling through the 3 plates,which are clamped together. Once the first hole expansion is achieved, a temporary boltis inserted around the sleeve, which is left in the hole to preserve correct alignment. Thenthe second hole is expanded with the same clamping system. The process is presentedschematically in Fig. 1.

Fig. 1. Stacked cold working process

Cold Working Process on Hard Metal Stacked Assembly 43

During the classic process, plates are expanded individually and the local deforma-tions of the metal plates create a volcano on each face of the coupon. In his works, [8]performed cold working expansion on several similar coupons and the measured heightsof such volcanoes were around a tenth of a millimeter. Although the coupons are stackedand strongly clamped in the present work, similar deformations are present, as shownby Fig. 2.

Fig. 2. Volcano formation at hole edge

3 Experimental Campaign

Assemblies in this study were made of titanium Ti-6Al-4V. Dimensions are given inTable 1, in accordance with the design in Fig. 3. Two different sizes of fasteners weretested, code-4 (6.35 mm diameter) and code-6 (9.352 mm diameter). Non-expandedassemblies were also tested for each code as references. Hard metals are usually stronglyloaded so it was interesting to study their behavior at high cold expansion rate. Thereforethe chosen DCE (degree of cold expansion) was 5% for each size. The coupons weredesigned to favor a net section failure at the first bolt of the middle part (Fig. 3).

The last step before the test was final assembly with the dedicated fasteners. In thisstudy, the bolts used were shear screws with self-locking nuts. Since T. Benhaddou [9,10] has demonstrated the major impact that preload has on the fatigue life of boltedassemblies, it was decided to use a high tightening torque to compensate the clearancebetween plates caused by the stacked expansion. The tightening torque Co was 29 N.mfor code-6 assemblies and 9.5 N.m for code-4. This tightening was executed with anumerical torque wrench having a precision range of 4%. However, strong uncertaintyremained between the torque applied and the existing preload in the fasteners due tothe numerous parameters involved in the process, such as torqueing speed, temperature,friction coefficients.

The fatigue tests were conducted with a frequency of 5 Hz and a load factor R= 0.1(see Eq. 2).

R = σmin

σmax= 0.1 (2)

The stress levels were chosen in order to obtain points ranging from 103 to 2.106

cycles. To take proper account of the dispersion effect inherent in fatigue tests, threeassemblies were tested for each configuration (size and DCE).

44 G. Pichon et al.

Table 1. Specimen designations and dimensions

Reference Designation Dimensions

Code T TTH DCE Ltot Ttot W Di Dr S

[−] [mm] [NA] [%] [mm] [mm] [mm] [mm] [mm] [mm2]

4-3-0-0 4 3.18 0 0 254 9.525 25.4 6.35 ± 0.02 6.35 60.48

4-3-0-5 4 5 5.93 ± 0.02

6-4-0-0 6 4.76 0 381 14.287 38.1 9.52 ± 0.02 9.525 136.09

6-4-0-5 6 5 8.89 ± 0.02

Fig. 3. Definition of assemblies

Specific metrology devices were installed to observe fretting behavior during thetests.

4 Results

The direct outputs of this experimental phase were the S-N curves or Whöler curvespresented in Fig. 4. A glance at the curves shows that the fatigue lives of the expandedassemblies (green) were lower than the references (blue and red).

Fig. 4. S-N curves a) Code-4, b) Code-6

Cold Working Process on Hard Metal Stacked Assembly 45

To explain such a difference, the failures of the assemblies were analyzed. Forexpanded assemblies, 100% of failures occurred in the net section as expected whereas,for the reference ones, only 95% of failures were located in the net section. The otherswere tangent to the hole. This failure mode can be explained by the strong preload ofthe fasteners.

Figure 5 shows the fretting zone of the assemblies tested and compares expandedand reference coupons for similar loading. On expanded assemblies, the fretting is con-centrated around the holes, while the fretting zones on reference coupons are larger andscattered all over the contact surface. From these observations, it can be deduced thatthe load transmitted by friction was significantly lower on expanded assemblies, so theload was mainly transmitted by the fasteners and the holes. This behavior is consistentwith the performance loss, assessed at about 15% (for 105 cycles).

Fig. 5. Fretting zones of expanded (blue) and reference (red) assemblies, a) Code-4, b) Code-6

The thermal evolution of the assemblies (reference and expanded, code-6, at 500MPaloading) is presented in Fig. 6. The reference assembly reaches a much higher tempera-ture than the expanded one during the first phase. More friction occurs in the referenceassembly, which is coherent with the previous analysis. Then fretting occurs and theheat created by friction decreases as more and more fretting occurs (appearance of blackpowder). The last peak of the curves shows the heat created during the plasticization ofthematerial. Once again, it can be seen that the reference peak ismuch higher than that ofthe expanded assembly. Plastic deformation is less pronounced on expanded assembliesas the cold working process reduces their plasticizing range.

46 G. Pichon et al.

Fig. 6. Thermal evolution of the assemblies during the fatigue test (Code-6)

5 Conclusion and Perspectives

Although this study has shown that the stacked cold expansion process combined withstrong tightening of bolts has a strong detrimental impact on the fatigue life of hard alloyassemblies, cold expansion alone is not to blame for such a performance decrease.

Deeper investigations should be conducted to determine the impact of cold workingon titanium bolted joints, with the implementation of different parameters and configu-rations. In the near future, experimental tests should be carried out with interface sealantto modify the fretting behavior, while others should be performed with hand tightenedbolts to remove the effects of bolt preload.

Simulations are also being developed to study the stacked cold expansion phe-nomenon, in particular the formation of volcanoes, which are at the origin of gapsbetween plates and the limited surface friction.

A scale effect was also observed in both reference cases, and expanded assembliesshowed a longer life in code 4 than in code 6. This scale effect is to be confirmed byother tests with larger dimensions.

Acknowledgments. The research work reported here was made possible by Airbus Operations.

References

1. Luchtvaart Nederland. Slim én duurzaam. Action plan. Ministry of Infrastructure,Amstelveen, The Netherlands (2018)

2. Huggins, G.L., West, R.R., Briscoe, R.T., Welch, J.M.: Engine pylon made from compositematerial (2012)

3. Boyer, R.: An overview on the use of titanium in the aerospace industry. Mater. Sci. Eng. A213, 103–114 (1996)

4. FTI. Cold expansion of holes using the standard split sleeve system and countersink coldexpansion. Process Specification 8101C. Fatigue Technology Inc. Seattle, WA (2014)

5. Quincey, D.E., Copple, C.M., Walsh, W.B., Jarzebowicz, R.Z., Easterbrook, E.T.: Split sleevecold expansion (1994)

Cold Working Process on Hard Metal Stacked Assembly 47

6. McCLUNG, R.C.: A literature survey on the stability and significance of residual stressesduring fatigue. Fatigue Fract. Eng.Mater. Struct. 30, 173–205 (2007). https://doi.org/10.1111/j.1460-2695.2007.01102.x

7. Leon, A.: Benefits of split mandrel coldworking. Int. J. Fatigue 20, 1–8 (1998). https://doi.org/10.1016/S0142-1123(97)00059-5

8. Achard, V., Daidié, A., Paredes, M., Chirol, C.: Optimization of the cold expansion processfor titanium holes. Adv. Eng. Mater. 19 (2016). https://doi.org/10.1002/adem.201500626

9. Benhaddou, T., Stephan, P., Daidié, A., Chirol, C., Tuery, J.-B.: Effect of axial preloadon double-lap bolted joints: numerical study, nantes. In: Proceedings of ASME 2012, 11thBiennial Conference on Engineering Systems Design and Analysis (2012)

10. Benhaddou, T., Daidié, A., Stephan, P., Chirol, C., Tuery. J.-B.: Optimization of fatiguebehavior of metallic shear joints. In: Proceedings of 12th International Conference on theMechanical Behavior of Materials - ICM12. Karlsruhe, Germany (2015)

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,adaptation, distribution and reproduction in anymedium or format, as long as you give appropriatecredit to the original author(s) and the source, provide a link to the Creative Commons license andindicate if changes were made.

The images or other third party material in this chapter are included in the chapter’s CreativeCommons license, unless indicated otherwise in a credit line to the material. If material is notincluded in the chapter’s Creative Commons license and your intended use is not permitted bystatutory regulation or exceeds the permitted use, you will need to obtain permission directly fromthe copyright holder.